twitter r t under crisis
TRANSCRIPT
Rumors vs Confirmed News on Twitter
Twitter Under Crisis-Can we Trust what we re-Tweet?
Clement Robert Kalthoom
Introduction
● Twitter is a micro-blogging service ● brings together millions of users ● enables real-time propagation of information to a large
group of users● not only enables the effective broadcasting of valid
news, but also of baseless rumors.
Chilean earthquake● The Chilean earthquake on Saturday , February 27, 2010 at 06:34:14 UTC
(03:34:14 local time)● It reached a magnitude of 8.8 on the Richter scale and lasted for 90
seconds.● it is considered the seventh stronger earthquake ever recorded in history.
studies over post-quake
● Research questions: To analyze the impact of Twitter on the propagation
of information during the Chilean earthquake,
● perform two types of studies over post-quake tweet data:
○ characterize the usage and social networks of the days immediately
after the event. The goal of this task is to observe how rumors and
news are propagated and the dynamics of the followers/followees
relationship.
○ investigate the ability of the social network to discriminate between
false rumors and confirmed news. To do this we examine tweets related
to confirmed news and to rumors, classifying manually each tweet.
Twitter Network During an Emergency
● Experimental Framework○ Collection of Users Activities
(Feb 27-March 2, 2010)■ Tweets: Tweets with
keywords related to the event.
■ Users Locations: Users around Santiago Timezone
○ Tweets: 4,727,524 tweets: 19.8% are replies to others
Twitter Network During an Emergency
● The Social Network
○ # of users: 716,344
○ #Following>#Followers(49.6%)
○ #Following<#Followers(46.2%)
● Authority Users with>100k
followers= 633
○ Most of them are
politicians, Mass Media &
celebrities.
Twitter Network During an Emergency
● # of tweets per User
○ >50% have 1 tweet
○ 11.47% of users have >10
tweets each
● The Average # of tweets (6.5) >
the Median
○ Outlier→ Some were tweeting more
than expected
Twitter Network During an Emergency
● Influence of Authority on # of tweets
produced
○ When the # of tweets decreases, the # of
followers decreases
○ 50 top users have high value of
followers
○ Top users are following less # of users
○ Top users: Celebrities, Politicians,
Mass Media,etc
Twitter Network During an Emergency
● Relationship between Top Users
○ CNNBBreakingNews does not follow anyone in top-20 users
○ Mass Media + NGOs + Individuals
○ Active users are strongly connected
Twitter Activity During an Emergency
● Comparison of 2 Trends during
○ Music Festival vs Earthquake○ Festival trend disappeared 20
minutes after the event○ Nbr of the Earthquake increased
since then.
Vocabularies used during the event per day.
Differ depending on the social needs.
Twitter Activity During an Emergency
● Re-Tweets Propagation
○ Retweet activity shows
how social network helps
in information
propagation.
○ How deep re-tweets cover
the social graph
indicates the relevance
of the tweet for the
community.
FALSE RUMOR PROPAGATION
● Veracity of information on Twitter and how this information is spread through the social network
● Manually selected some relevant cases of valid news items, which were confirmed at some point by reliable sources.
● Manually selected important cases of baseless rumors which emerged during the crisis (confirmed to be false at some point).
Classification results for cases studied of confirmed truths and false rumors
1. Goal is to observe if users interact in a different manner when faced with these types of information.
2. Each case studied was selected according to the following criteria:
a. Asignificant volume of tweets is related to the case(close to 1,000 or more).
b. Reliable sources (external to Twitter) allow to asses if the claim is true or false.
Classification of results for cases studied of confirmed truths and false rumors● These results show that the propagation of tweets that correspond to rumors differs
from tweets that spread news because rumors tend to be questioned more than news by
the Twitter community.
● This fact suggests that the Twitter community works like a collaborative filter of
information.
● This result suggests also a very promising research
line:
○ it could possible to detect rumors
by using aggregate analysis on tweets?
Conclusion and Lesson Learnt● +
○ Social Networks Analysis revealed interesting facts:■ Rumors are those tweets with Many Questions and Denies■ Confirmed News get Many Affirms.
■ Twitter Community works like a Collaborative filter of information
● - ○ No automatic way to classify confirmed news against false rumors
■ If it was there, would it be able of analysing facts like sentiments, sarcasm etc?
■ And Label tweets as Questionable or Not.○ Social Side of the Issue vs Social Community on Twitter
■ Who’s in a position to confirm news?
● Tsunami was real but, Government officials had denied its existence!
Thank you